Formality and Informality: Lessons from the New Findex Survey
The Findex project helps to correct a long-standing imbalance in evidence on global finance: an abundance of data on the supply of financial services but curiously little that’s systematic and comparative about global demand. Together with the IMF’s Financial Access Survey, we’re finally getting a clear picture of the holes in global financial access.
There’s a lot to celebrate now that the Findex is here. So much so that it’s striking that it took so long to create a constituency for the efforts. Stanley Fischer had initiated the push in 2004 as head of the Advisors Group for the UN Year of Microcredit, and I joined Princess Maxima of the Netherlands in pushing the agenda forward in advisory roles with the UN in 2005. But it was the Bill and Melinda Gates Foundation’s support of the World Bank’s Development Economics Vice Presidency (DEC) in 2010 that ultimately got us here.
The Findex headline turns out to replicate an earlier count of the global financial gap: half the world is unbanked, about 2.5 billion adults, the same bottom line that came from aggregating a range of independent surveys from ten years ago. But even if the headline is familiar, the Findex delivers rich details.
A few years ago, Daryl Collins, Stuart Rutherford, and Orlanda Ruthven tackled the demand side by going deep rather than wide. In the research which came together as Portfolios of the Poor, about 250 households in three countries were interviewed intensively over the course of a year. While Portfolios boasts the sub-title “How the World’s Poor Live on $2 a Day,” it was always a mighty stretch to extrapolate from 250 households to 2.6 billion people living on under $2 a Day. While the Portfolios of the Poor research involved intense engagements (meetings with the families every two weeks), the Gallup survey on which the Findex draws involves one-time conversations, often by telephone. The compensation is that the Findex reports on over 150,000 people in 148 countries.
So how do Portfolios of the Poor and the Findex line up, especially on informal finance?
First, there are some important areas of agreement. Two stand out for me. One is that informal savings mechanisms are important for savers, even among those who also have formal accounts. The informal mechanisms include savings clubs with neighbors and families, such as ROSCAs and ASCAs. In keeping with the spirit of the Portfolios research, roughly half of the Findex adults who saved using an informal mechanism also saved in a formal financial account. Those informal mechanisms turn out to be convenient and provide structure that help keep temptation at bay. On the other hand, they’re not as reliable as formal accounts.
What does this mean for policymakers? First, simply providing access to savings accounts won’t necessarily be enough to spur their active use. Second, as new work in behavioral economics shows (e.g., work on commitment savings contracts), success can be found by creating formal account mechanisms that adapt ideas from the informal.
A second important place where Portfolios of the Poor and the Findex data line up is with informal borrowing. Here, the important shared take-away is the overwhelming importance of friends and family as a source of informal credit. The flip side of that is the relative lack of importance of informal lenders (at least relative to what’s suggested when reading the microfinance literature).
For years, microfinance rhetoric has wrongly equated informal credit with moneylenders charging exorbitant interest rates. That equation is often deployed to justify the high interest rates charged by microcredit institutions (“Well, if they didn’t use microcredit, customers would have to pay the moneylender over 100% per year…”). The trouble with the argument is that going to the moneylender is not the only substitute for microcredit borrowing. Loans from family and friends are not only prevalent, but they usually come with no interest charges at all (though they may carry the costs of reciprocal obligations and the like). Portfolios of the Poor argues the case on the basis of data 250 households, and now the Findex shows a similar pattern for 150,000:
These data also help to make sense of the finding that microcredit customers do in fact care about interest rates.
Portfolios of the Poor and the Findex data most notably diverge when it comes to the intensity of financial activity. Portfolios showed extremely active borrowing and saving over the course of a year, much of it small scale and short term and almost all informal. That same level of intensity is not nearly as clear in the Findex data. Demirgüç-Kunt and Klapper find regional average savings levels of 40% at best (this includes saving via informal savings mechanisms):
Why don’t we see more saving? One reason is that the informal savings mechanisms are narrowly defined. The question on the survey only asked about setting aside money “using an informal savings club or a person outside the family.” The question is the right one to ask in order to get a precise answer to a precise question – and it’s much better than the alternative of getting an imprecise answer to an imprecise question. Still, we have to be careful in interpretation. A lot of people save by simply holding onto cash, either keeping it on their person or hiding it at home. Others save in ways that they might not think of as saving. Making a loan to a friend is one example (as the data show, those kinds of informal loans are quite common).
A second issue is that people evade questions and forget activities – or never know about their financial holdings in the first place. Dean Karlan and Jonathan Zinman find that there’s a lot of lying about (really expensive) consumer borrowing in South Africa. Daryl Collins finds that people under-report insurance and pensions (also in South Africa) but not saving or borrowing, often because they don’t know that their employer is facilitating insurance and pensions via paycheck deductions. Robert Cull and Kinnon Scott find that it matters who you ask and how you ask. Household heads turn out to do a much better job of reflecting the household’s financial activity than other adults. The Findex data, by contrast, asks questions of a randomly chosen adult, not necessarily the household head. Portfolios of the Poor shows that constructing the full story takes much more than a single interview.
The bottom line is that we can’t take all of the Findex data at face value. There’s likely a lot of under-counting, especially of informal mechanisms. Still, the Findex data show the broad shape of financial access for the first time using a unified approach. That’s a powerful start.